摘要： 传统本体演化方法都是顺序地处理多个本体演化需求，其中关于演化代价的影响因素不明确。针对该问题，分析影响演化代价的因素，提出一种基于上下文窗口的本体演化方法，将演化过程转变为图的启发式搜索过程，结合窗口内信息，从局部范围内全局地选择演化路径。实验结果表明，与传统方法相比，该方法能够进一步降低演化代价。

关键词: 本体演化, 演化代价, 演化路径, 上下文窗口, 演化算法

Abstract: Approaches of ontology evolution are deal with evolution requirements in sequence, the factors of evolution cost are uncertainty. To solve these problems, it redefines the approach of computing evolution cost, and proposes an ontology evolution approach based on context window. It transforms an ontology evolution process into a heuristic graph searching process, combines the inner information of context window, and selects an evolution path from a full view in the range of context window. Experimental results show the approach can decrease the evolution cost than traditional approaches.

Key words: ontology evolution, evolution cost, evolution path, context window, evolution algorithm

中图分类号: 

• N945